AI continues to take the tech community, and the world, by storm
For the majority of people, it wasn’t long ago that Artificial Intelligence was synonymous with digital personal assistants like Siri and Cortana. AI research was kept largely under the radar, and confined to big technology companies. 2017 proved this has never been further from the case. AI is now a business necessity when it comes to understanding and analysing valuable data. Aside from dealing with data, AI is now used by companies across the spectrum to automate processes, detect fraud, respond to customer queries and carry out menial, but important, administrative tasks. Artificial Intelligence is only improving. What were the most important developments of 2017, and how will they shape the future of AI?
1. Swarm AI predicts TIME’s Person of the Year
TIME Magazine’s 90 year tradition of naming a Person of the Year. The winner isn’t necessarily everyone’s favourite candidate, but rather the individual (or, as the case may be, the social movement) that has most influenced the news. In 2016, San Francisco based AI company Unanimous AI correctly predicted that Donald Trump would win. Last year, they did it again, successfully naming the #MeToo campaign as the winner. Unanimous AI used Swarm AI technology to ask 60 members of the public who was least likely to win in each round, whittling down the 33 potential candidates by process of elimination. AI improves over time, and therefore so will the accuracy of its predictions.
2. AI changes employer needs
On the topic of enhancing human performance with AI, many companies are coming to view Artificial Intelligence as a creator of, rather than a threat, to jobs. One of the main barriers to the adoption of AI is the belief that it will lead to mass unemployment. Some jobs – 861,000 in the UK alone – will be replaced by automated technology. Despite this, there has been a growing recognition that AI also creates employment opportunities that weren’t there before. Employers are increasingly on the lookout for coders, programmers and technicians to monitor and maintain complex, artificially intelligent systems.
“The overall global trend is that when you implement new technology, over time you actually create more jobs,” says Claus Risager, CEO of Blue Ocean Robotics. According to Claus, there are more global jobs and working hours per day than ever before.
3. AI as-a-service
It will come as no surprise that AI has been integrated into the as-a-service business model. This makes resources (namely infrastructure, platforms, and software) available to users via the internet. 2017 was the year that Software-as-a-Service (SaaS) matured, driven by Google, Microsoft and Salesforce. The as-a-service model is enabled by cloud computing, but in order to handle an influx in data and demand, the cloud has needed to get smarter. Providers turned to deep learning, a subfield of machine learning, to improve efficiency. The wider use of AI tools has and will continue to be encouraged by the adoption of cloud computing.
4. AI moves into the home
If you own an Amazon Echo or Google Home device, you’ve invited AI over the threshold and into your domestic life. AI is the technology that enables them to answer questions, give recommendations, provide entertainment, control smart homes and more. Recognising a user’s identity via advanced voice recognition, combined with user preferences, has enhanced marketing techniques and uncovered yet another application for AI. Last year, Google Assistant went head to head with Amazon Alexa, and Apple announced the development of a competing model called HomePod in June. Each of these products have increased consumer familiarity with conversational interfaces powered by Artificial Intelligence. Ordinary people as well as corporations can now use AI to improve their daily lives.
5. AI creates AI
Sending data through neural networks with deep learning techniques is notoriously time consuming and complicated. Google’s answer was to design an artificially intelligent system that could do this itself. In May 2017, the company created AutoML. AutoML consequently built a daughter AI called NASNet to recognise objects in real time video. It did so with 82.7 per cent accuracy, making it the most effective image recognition tool in existence. In other words, AutoML is better at making AI than we are. Combine this with AI’s ability to teach itself, and the singularity could be closer than we think. At the same time, AI-made AI ties into the theme of making the technology more widely available, as users could use AutoML to build AI for specific tasks without needing to have a background in coding.
Given the benefits of AI, the technology’s explosion is unsurprising. AI’s usefulness for business has been reflected in the sheer number of companies that now use it in some form. It’s not difficult to imagine that one day, businesses may be unable to function optimally without it. The changing attitudes towards implications for employment have also been instrumental in encouraging adoption. Nonetheless, there is ongoing unease at the prospect of the singularity – the point in time at which Artificial Intelligence exceeds human intelligence. Now that AI can create AI, the creation of a code of ethics for AI’s continued development is vital.